HemaYugma

HemaYugma - Smarter Matches, Safer Lives

Comment

💡 Project Story: HemaYugma

Smarter Matches. Safer Lives.

🚀 About the Project

HemaYugma is an AI-powered platform that reimagines blood compatibility for Thalassemia patients by going beyond ABO and Rh — it considers genetic-level antigen matching to personalize every transfusion.

As a solo participant, I wanted to tackle a real, underexplored medical issue using technology — something with both social impact and scientific complexity. This led me to a startling realization:

Even "compatible" blood can harm if minor antigens don’t align.

That's the challenge I chose to solve — and HemaYugma was born.

🎯 Inspiration

While exploring ways to support the Blood Warriors initiative, I came across cases where Thalassemia patients suffered immune complications even with matched blood types. These issues stemmed from alloimmunization caused by minor antigen mismatches.

One sentence in a medical forum changed everything:

"We’ve had to tell patients they couldn’t be transfused — not because there was no blood, but because they’d developed antibodies to all compatible types."

That broke my heart. And it sparked my mission:
Create a system that understands the patient's immune profile and predicts donor safety.

🧠 What I Learned

This project challenged me to dive into fields beyond my core — including:

  • Hematology fundamentals: Antigen systems like Duffy, Kidd, Kell
  • Genomic data interpretation: Allele matching, WGS basics
  • Privacy-by-design in healthcare platforms
  • Designing for ethical AI in medicine

And perhaps most importantly:

How to design alone — yet build something meant for everyone.

🛠️ How I Built It

As a solo builder, I leaned on modular tools and layered architecture.

🧱 Stack Overview:

Layer Stack Used
Backend Java Spring Boot, PostgreSQL, JWT Auth
AI Engine Python (Scikit-learn, Pandas)
Frontend React.js (web), mockup only
Security AES-256 encryption, secure REST APIs
Integration Mock e-RaktKosh API + location services

🩸 Key Features:

  • Upload patient DNA/antigen profile
  • Predict safe donor matches using ML
  • Generate Blood Compatibility Passport
  • Visual score + donor alert system
  • Modular design for future hospital integration

⚙️ Challenges I Faced

🔐 Data Sensitivity

Handling patient DNA requires maximum data protection. I implemented encryption, access control, and local-only processing.

🧪 Dataset Limitations

India lacks minor antigen-level donor data. So I built a prediction engine based on:

  • Simulated profiles
  • Ethnicity inference
  • Donation history patterns

⏱️ Solo Development

Building a high-impact health-tech prototype alone in 48 hours tested my time management, research depth, and stack decisions.

🌈 What’s Next

Post-hackathon, I plan to:

  • Collaborate with Thalassemia NGOs to run a pilot
  • Improve donor-side input with a progressive form flow
  • Launch HemaYugma as an open-source project to bring blood matching into the AI age

🧬 In Summary

HemaYugma bridges science and compassion — using AI to protect lives at the genetic level. Built solo with heart, code, and a vision, it transforms how we match donors to patients.

Because when lives depend on a match...

Smarter Matches. Safer Lives.

Built With

Share this project:

Updates

Submission history